Brief Introduction

After two decades of remarkable stability, Bolivia’s democratic future suddenly became uncertain when, on 17 October 2003, Gonzalo Sanchez de Lozada resigned as president in the face of month-long civil unrest. Bolivia suddenly joined the ranks of other South American countries whose presidents had been forced from office by a popular uprising. But for the two previous decades, Bolivia’s political system had been marked by a degree of political stability uncommon in the Central Andes. This stability was often credited to the country’s unique political institutions, labeled “parliamentarized presidentialism” by Rene Antonio Mayorga. This study asks two questions: A) did the institutions of parliamentarized presidentialism facilitate the previous two decades of political stability and B) if so, how can we explain the events of the 2003 “guerra del gas”?

This study defines “political stability” in a minimal sense: the absence of coups or popular uprisings, and evidence that all major players accept the procedural democratic norms. I do not argue that Bolivia’s democracy was consolidated, only that it had passed democratization’s transition phase. While stability is an essential component of democratic consolidation, most authors include other factors such as transparency and the rule of law in defining “consolidation” – elements Bolivia clearly lacked.

Others have pointed out Bolivia’s unique hybrid political system. Thus, this study pursues the following research hypothesis:

Stability as a consequence of political institutions that gave political elites incentives to pursue cooperative bargaining strategies.

Like other South American countries, Bolivia is nominally a presidential system. Unlike its neighbors, however, Bolivia’s electoral system combines executive and legislative elections in unified proportional representation lists. And in the absence of a clear majority winner, Bolivia’s executive is elected by the legislature, not by a run-off election. Proponents of “parliamentarized presidentialism” argue that this unique institutional design provides strong incentives for political elites to develop cooperative bargaining strategies similar to those employed in traditional parliamentary systems. Like most parliamentary systems, Bolivian governments are multiparty coalitions developed after intense negotiations during the legislative run-off.

If “parliamentarized presidentialism” explains Bolivia’s previous political stability, then how can we account for the political system’s breakdown in October 2003? Is it possible that other factors facilitated stability in the 1982-2002 period? This study pursues three alternative hypotheses.

1. Stability reinforced by a common national consensus on the 1952 National Revolution. Since most political actors in the post-transition period saw themselves as descendants of the 1952 Revolution, political stability may have been supported by a lack of significantly divisive ideological cleavages or contradictory definitions of the “nation”.


2. Stability reinforced by consensus on neoliberal economics. After the failed 1982-1985 Siles Zuazo government, all major political parties broadly supported neoliberal economic policies, even if there was disagreement on specific implementation.


3. Stability reinforced by political clientelism. It is also possible that Bolivia’s political system was reinforced by corruption and spoils-sharing between elite-led political parties.

Of course, it is possible that any combination or all of these factors combined to different degrees to facilitate political stability at the national level. The breakdown of the status quo, therefore, should be explained by a change in any of the four explanatory variables.

This study also examines whether changes in political institutions affected the ability of elites to build cooperative bargaining strategies. In 1994, amendments to Bolivia’s constitution changed the electoral system from list proportional representation (List-PR) to multimember mixed-member proportional representation (MMP). The change allowed voters to elect half the lower house from single-member, plurality-winner districts, with the other half elected by awarding compensatory seats from department-wide party lists; senators were still elected from single party lists. Recent evidence suggests the change encouraged political parties to pursue regional, rather than national, agendas, cementing themselves in regional power bases. This study therefore also tests whether the change from List-PR to MMP encouraged the formation of political cleavages, which may have led to a breakdown of elite- and social-level consensus.

To test these hypotheses, this study relies primarily on three types of data: A) electoral data, B) content analysis of elite political discourse during elections, and C) cabinet-level data.

Using electoral data, this study analyzes shifts in voter behavior across five national elections (1985, 1989, 1993, 1997, 2002). The focus is on establishing voter trends – or an “electoral geography” – that accounts for party strengths across regions, examining how electoral alignments changed between elections. These trend data analysis will consider variables such as: electoral volatility, vote-to-seat disproportionality, change in vote share for the three “traditional” political parties (MNR, MIR, ADN), and change in vote share for “incumbent” parties (that is, parties belonging to the incumbent government coalition). This data seeks to demonstrate that the degree to which the traditional parties lost support varied across regions, underscoring that Bolivian politics since the change to MMP has become more regionally-oriented, centering on growing regional cleavages.

This study also analyzes elite campaign discourse as presented in national newspapers for all five national elections. Using content analysis, this study seeks to discern whether political elites ever shared consensus along basic issues by dissecting elite campaign rhetoric. Statements are broken down into two basic areas: A) economics and B) national identity. The analysis seeks to determine whether elites in general – or elites from specific parties – tended to agree on basic issues related to national economic orientation and/or tended to appeal to a common view of Bolivian national identity. The question of a common view of the “nationality” question is important as contrast to the present regionalism in Bolivian politics; it is important to understand whether regionalist politics is a new phenomenon or not.

Content analysis data is used to test whether political elites “learned” to operate under institutional incentives. That is, analysis of elite discourse seeks to find evidence of inter-party “signals” sent to other elites that demonstrate a willingness to engage or not in cooperative bargaining strategies with specific other actors. If the institutions of parliamentarized presidentialism shaped elite strategies, then the number of strategic “signals” should increase from election to election. The study also seeks to determine whether cooperative “signals” increased or decreased after the change to MMP, as well as their relation to changes in other dimensions of political discourse (i.e. economic or national identity).

Finally, this study looks at longitudinal cabinet-level data for the 1982-2003 period. This allows for an analysis of coalition behavior between elections, primarily to test whether cabinet stability – a surrogate for coalition stability – changed over time. If elite consensus became increasingly difficult over time, we should see increasing difficulties in maintaining a stable distribution of cabinet positions between coalition members. The focus here is less on changes in cabinet personnel, but rather on changes in partisan composition of cabinets.

The October 2003 “guerra del gas” highlighted deep political cleavages between Bolivia’s regions – specifically between the Andean and lowland departments. This study pursues the secondary hypothesis that the change to MMP increased regional political schisms by offering elites incentives to appeal primarily to local issue, rather than national politics. Such findings have important consequences for the study of democratic institutions, especially the recent trends in “constitutional engineering”. If the transition to MMP was a contributing cause to the breakdown of elite consensus and political stability, then it poses serious questions about the desirability of such an electoral system in other new democracies.